AN EVALUATION OF MEASUREMENT OF ADHERENCE AND PATIENT RECRUITMENT METHODS IN PATIENT ADHERENCE TO MEDICATION RESEARCH

Abstract

<p><strong>Objective: </strong>To provide an overview of the state of trial methodology concerning measurement of patient adherence and patient recruitment, and explore how the quality of these methods impact the adherence results found in trials.</p> <p><strong>Data sources: </strong>Major bibliographic sources, reference lists, and clinicaltrials.gov were searched for relevant trials up to January 2013.</p> <p><strong>Study selection: </strong>Approximately 150 trials were included in the full systematic review, from which 50 trials were selected to represent several methods of measuring adherence.</p> <p><strong>Results: </strong>There were a variety of different measures of adherence with qualities ranging from valid and unobtrusive, to unreliable and subjective. The median overall quality of adherence measures was 5 (IQR 3, maximum score 9, higher is better). The overall correlation of the quality of the measures of adherence and the coefficient of variation (CV) or proportion adherence suggested that adherence measures rated as higher quality were associated with a higher CV but not associated with a lower proportion adherence. The median overall quality of patient recruitment methods was 2 (IQR 1, maximum score 6, higher is better). The correlation of the power of a trial and the quality of the patient recruitment methods, was slightly positively correlated for both binary and continuous data.</p> <p><strong>Conclusions: </strong>The quality of methods employed in adherence trials varies considerably and affects at least some findings of these trials. The importance of these differences in quality merits further study, but it is clear that better standards of adherence measurement are needed to support adherence research.</p>